Per configuration
Report 21/06/2021
¶
Questions
¶
Once noisy data is "cleaned" in first iteration then only use minimal renoising ($0\% \to 10\%$) for better training, i.e. simulates noisy input cable but then perfect "iterative" system. Not much learning from second iteration
Training structure
¶
Loss
¶
Huber loss
with $\delta =1$
Metrics
¶
Huber loss
MNIST classifier (2 layers 32, 10 dense)
Noise visualization
¶
Noise level mean comparissons
¶
Huber loss
Classifier Accuracy
Detailed results per experiment
¶
Noise with 0.0% to 0.1%
Noise with 0.2% to 0.3%
Noise with 0.4% to 0.5%
Noise with 0.6% to 0.7%
Noise with 0.8% to 0.9%